This project will identify mechanisms underlying age-related changes in cognition. Two general hypotheses as to the specific nature of these mechanisms will guide the research. Under one hypothesis, inhibitory processes become less effective in older adults, while under the other, age weakens the strength of connections between the units of memory, so that activity transmitted from one unit to another declines with age. Empirical support for one or the other hypothesis that will emerge from the proposed research promises to have a major impact on the field. By examining a series of phenomena that have rarely if ever been examined in older adults, e.g., repetition blindness, illusory conjunctions, effects of delayed and amplified auditory feedback and the relative ability to perceive vs. produce speech, the project promises to fill several gaps in our knowledge. Moreover, the project focuses on tasks and phenomena where older adults are predicted to exhibit superior performance to younger adults, unlike previous work where declines in the abilities of older adults have been the main focus. The project also carries theoretical and practical implications for many areas of cognitive psychology, including seemingly unrelated areas such as word retrieval failures, and the etiology of stuttering. Finally, the project will develop a testable and well specified model of aging and cognition that is computational in nature. Five sets of studies are proposed. The first two sets compare the perceptual errors of young and older subjects in detecting rapidly presented letters and acoustically presented words. A third set examines the effects of aging on experimentally induced errors in production. A fourth set compares the ability to comprehend vs. produce speech in young and older adults to determine why production processes tend to become impaired with aging, whereas comprehension processes remain relatively intact. The fifth set of studies will develop computer simulations of how cognitive processes change with age.